Abstract— The use of data-envelopment analysis (DEA) to determine the most efficient decision making unit (DMU) has recently drawn attention in the literature. For some applications of DEA, decision-makers may only want to identify the most efficient DMU rather than determining the efficiencies of all possible DMUs. Some recent approaches combine the use of DEA and discriminant analysis (DA) to rank DMUs and identify the most efficient DMU. However, some of these approaches have drawbacks. This paper addresses those drawbacks and offers suggestions for improvements. This paper introduces a modified model to solve the problem of multiple solutions in DEA in determining the most efficient DMU. The modified model has a new goal, based on the assumption of cluster analysis that objects belonging to the same group should be more similar to each other than to objects from other groups. With this new goal, the modified model selects the solution in which members of the same group, whether efficient or inefficient, are most tightly clustered.